284 CHEMICAL ENGINEERING
rectangular tank the velocity of approach is the velocity in the body of the tank. The rectangular tank shows a greater depth of water at the discharge than the circular one because the overflow is more restricted and based on cubical contents; the equal settling capacity of the two shape of tanks is closer. One suspects that a complete theory would show that tanks of equal cubical contents would have equal settling capacity.
As a practical settling device the circular tank will be superior to the rectangular tank because of the quieter condition at the overflow, there being very much shallower overflow and consequently less tendency for converging currents to sweep partially settled material out of the tank. It is this quiet action at the periphery which gives its great advantage in settling material so finely divided as to be in the colloidal state or approaching the degree of subdivision that the phenomena of that state implies. If by the time the very finest particle reaches the periphery it is below the shallow depth of the overflow current then regardless of its rate of settlement it will eventually reach the bottom of the tank and can be discharged by means suitable for that purpose.
Capacity of Tanks on Very Finely Divided Matter.—The object of the separation between very fine matter and liquids is the production of a clear overflow and as great density as possible in the discharging thickened material. When relatively large proportions of finely divided matter and water are mixed and allowed to settle the mixture will be observed to arrange itself into zones or bands on account of a tendency of the particles to act rather as masses than individuals. The heaviest particles will settle promptly to the bottom of a glass test container while the balance will tend to arrange themselves in layers, each successive layer towards the top of the test vessel having ranges of grains with successively less average settling rate. As the zones settle clear water appears above the topmost one and the rate at which the surface of the topmost zone subsides is the practical rate of settlement for the material being tested. After a lengthy time it will be found that the subsidence of the upper surface of the top of the settling material practically ceases; when this condition is reached, the maximum density for thickened material has been reached and it is only necessary to withdraw some of the settled material and determine the proportion of dry solids to find o.ut the greatest density of pulp which may be withdrawn from the settling tank.
In a settling tank with continuous entry of fresh material there is confusion of the zones owing to the fact that particles and floes of material of comparatively rapid settling rate are settling through those of a slower rate of subsidence but as in the test vessel there is a lowermost zone whose boundary surface does not change so long as withdrawals are constant and there is no change in the character and quantity of the feeding stream. As must be evident there is a gradation in density measured from top to bottom of the settling tank ranging from that of the entering stream, or less, to that of the thickened material at the bottom.
In determining the capacity for Dorr thickeners the wet-crushed material is made up into a series of pulp mixtures with varying proportions of water and these are submitted to test for rate of settlement and the most practical maximum density of thickened material. The greater the proportion of water the faster the material will subside but on the other hand the greater the amount of clean water which must be removed, so that starting with a large proportion of water a large capacity will be shown owing to the greater influence of the higher settling rate and this will gradually diminish until a settling point is reached when the settling rate begins to show a stationary condition, that is, that the rate from one test to the next differs so little